An Efficient Evolutionary Algorithm for Multi-Objective Stochastic Job Shop Scheduling

This paper addresses multi-objective job shop scheduling problems with stochastic processing time. The objective is to simultaneously minimize the expected makespan and the expected total tardiness. A new permutation-based representation method is first proposed, in which the substring related to each machine is a permutation. The conflict is eliminated by giving priority to the operation with the minimum gene value among the conflicting operations in the same permutation. An efficient multi-objective evolutionary algorithm is then presented, which archive maintenance and fitness assignment are performed based on crowding measure. The proposed algorithm is finally applied to some benchmark problems and computational results demonstrate that the proposal algorithm has promising advantage in stochastic job shop scheduling.